Extended Data Fig. 7. MECOM down gene network enrichment is independently associated with overall and event-free survival.
(a) GSEA of MECOM down genes in primary AML patient samples from TCGA (left) and BEAT AML (right) stratified by MECOM expression. Individual gene expression was averaged from TCGA or BEAT AML samples with MECOM expression of log2(RPKM) > 4 and compared to the average of samples with MECOM expression log2(RPKM) < 4. The Kolmogorov Smirnov (K-S) test was used to determine the significance of GSEA. (b-d) GSEA of MECOM down genes in primary AML patient samples from TCGA. For each patient sample, expression of every gene was compared to its average expression from all TCGA patient samples, and GSEA was performed to assess for enrichment of MECOM down genes. Representative plots of three individual patients are shown. (b) Patient 2896 had enrichment of MECOM down genes and an overall survival of 230 days. (c) Patient 3011 had depletion of MECOM down genes and an overall survival of 2450 days. (d) Patient 2982 had no significant enrichment or depletion of MECOM down genes and an overall survival of 1110 days. The Kolmogorov Smirnov (K-S) test was used to determine the significance of GSEA. (e-f) Stacked bar graph showing proportion of patients with MECOM network enrichment or depletion following stratification by clinical risk group or LSC17 score in adult (e) or pediatric AML (f). (g-k) KM event-free survival curves for the pediatric AML cohort stratified by (g) MECOM expression, (h) MECOM network enrichment, (i) MECOM NES, (j) clinical risk group, and (k) LSC17. For continuous variables in (g), (i), and (k) the optimal threshold was determined by maximizing sensitivity and specificity on mortality (Youden’s J statistic). Hazard Ratios (HR) were computed from univariate cox-proportional hazard models. P values representing the result of Mantel-Cox log-rank testing are displayed. Test for trend was performed for clinical risk group stratification (>2 groups).